In previous studies, we have demonstrated a fibrocartilaginous membrane around hydroxyapatite-coated implants subjected to micromovement in contrast to the fibrous connective tissue which predominates around similarly loaded titanium alloy implants. In the present study, in mature dogs, we investigated the effect of immobilising titanium (Ti)- or hydroxyapatite (HA)-coated implants already surrounded by a movement-induced fibrous membrane and compared the results with those of similar implants in which continuous micromovement was allowed to continue. The implants were inserted in the medial femoral condyles of 14 dogs and subjected to 150 microns movements during each gait cycle. After four weeks (when a fibrous membrane had developed), half the implants were immobilised to prevent further micromovement. The dogs were killed at 16 weeks and the results were evaluated by push-out tests and histological analysis. The continuously loaded Ti-coated implants were surrounded by a fibrous membrane, whereas bridges of new bone anchored the HA-coated implants. The immobilised implants were surrounded by bone irrespective of the type of coating. Push-out tests of the continuously loaded implants showed better fixation of those with HA coating (p < 0.001). The immobilised Ti-coated implants had four times stronger fixation than did continuously loaded Ti-coated implants (p < 0.01) but there was no equivalent difference between the two groups of HA-coated implants. The amount of bone ingrowth was greater into immobilised HA-coated implants than into immobilised Ti-coated implants (p < 0.01). Two-thirds of the HA coating had been resorbed after 16 weeks of implantation, but 25% of this resorbed HA had been replaced by bone.(ABSTRACT TRUNCATED AT 250 WORDS)